Hall effect in pinned and sliding states of NbSe3

1
Hall effect in pinned and sliding states of NbSe3

• A. Sinchenko, R. Chernikov, A. Ivanov
• MEPhI, Moscow
• P. Monceau, Th. Crozes
• Institut Neel, CNRS, Grenoble

2
Outline

• Hall effect in CDW compounds. Motivation of
  studying of NbSe3.
• Hall effect in NbSe3 at low longitudinal electric
  field EltEt. Comparison with magnetoresistance.
  Two band model problems.
• Hall effect in NbSe3 in sliding state of CDW.
  Hole and electron pockets what is the
  difference?
• Conclusion

3
Hall effect in NbSe3 (EltEt)

• NbSe3 in the Pierlse state semimetal ground
  state because small electron
• and hole pockets in the Fermi surface
  main contribution to the Hall effect from
  pockets carriers
• Hall voltage is quite unusual in NbSe3 below
  Tp259 K
• - strong non-linear magnetic field dependence
  which drives the Hall voltage through a negative
  minimum and then positive at higher fields.
• Monceau and Ong (1978) reported that the
  zero-field-limit Hall constant changes sign from
  n-type to p-type at 15 K.
• Explanation in the frame of a two-band model (Ong
  1978) in which the difference in population
  (np-nn) 31018 cm-3 below Tp2.

What is correct?
R.V. Coleman, et al, PRB, 1990
4
Hall effect in the compounds with a CDW. (EgtEt)

• Does CDW give contribution to the Hall voltage?

TaS3 S. Artemenko, et al, JETP Lett., 1984
K0.3MoO3 L. Forro, et. al, PRB, 1986
Explanation CDW itself does not give
contribution to the Hall voltage. Non linear Hall
voltage is result of normal metal back-flow (S.
Artemenko and F. Kruglov, (Sov. Phys. Solid
State, 1984)
5
Hall effect in NbSe3 (EgtEt)

• NbSe3 G.X. Tesseme and N.P. Ong CDW motion
  gives no visible contribution to the Hall effect.
  (PRB, 1981)

Back-flow model predicts reduction of VH also
in NbSe3 especially at the temperatures close
to Tp.
Where is back-flow?
6
Experimental
Two types of structure

• Evaporation of gold contacts
• to the opposite faces of crystal

2. Preparation of Hall probes from the crystal
itself
In both cases the change in voltage on the Hall
pairs of contacts V1,3(B)-V1,3(-B)/2 or
V2,4(B)-V2,4(-B)/2 was taken to be equal to
the Hall voltage, VH, and the sum V1,3(B)V1,3(-
B)/2 or V2,4(B)V2,4(-B)/2 was taken as a
longitudinal drop of voltage.
7
Experimental results (EltEt)
At low electric field (EltltEt) the Hall voltage
is linear function of current. CDW does not give
the contribution to the electric transport.
RHVH/I
5K
35K
RH is strongly magnetic field dependent and
demonstrates the reversal of the Hall constant
sign at all temperatures
8
Comparison with magnetoresistance
maximum dR/dB at BB0
BltB0 MR B2 BgtB0 MR B
9
Comparison with magnetoresistance
B0Bzc ?
Electron and hole pockets demonstrate quite
different behavior in magnetic field electrons
MRB2 holes MRB
10
MRB2 - usual MRB - unusual
Quantum linear magnetoresistance ?
Conditions (Abrikosov, 1999)
In the case NbSe3
n 1017cm-3 and m 10-2me
Two band model n 1018cm-3 and m0.24me
1 high -T CDW 2 - low - T CDW 3 - ?
NbSe3 3 types of chains
heavy electrons
small hole pockets because non-perfect nesting
of low-T CDW (m 10-2me)
Correlation with ARPES data (J. Schafer et al,
PRL, 2003)
11
Experimental results (EgtEt)
30 K
A first step what is the Et(B)
dependence? Threshold electric field is
practically Independent on magnetic field
at Tgt25 K.
Experimental results above 25 K
12
30 K
Second type of Hall contacts (litography)
First type of Hall contacts (evaporated)
In both cases strong influence of CDW motion on
the Hall voltage.
Three possibility to explain
1. Field generation model CDW generates normal
carriers 2. CDW motion deforms electron or hole
pockets. 3. back-flow model if CDW interacts
with one type of carriers only (with light
holes?).
13
To make this effect more pronounce, we determine
the difference dVHVH-Vlin The
last term is the linear Hall voltage dependence
observed at low electric field below the
threshold.
30 K
Below Et visible deviation of the Hall voltage
from linear dependence. Position of this
deviations coincides with corresponding
singularities in IV-curve. Most probably this
effect is attributed by the local CDW
deformations.
14
Temperarure evolution
Deviation of the Hall voltage from linear
dependence decreases with the temperature
increase.
d VHAexp(-T/T0) T03.4 K
15
conclusion
1. Hall voltage, VH, changes sign in a wide
temperature range and the magnetic field for
which VH crosses zero is temperature
dependent. The two band model needs
corrections. 2. Electron and hole pockets
demonstrate qualitatively different
magnetoresistance behavior. 3. in high magnetic
field the CDW motion changes significantly the
Hall voltage at all temperatures below Tp2, that
can not be explained in the frame of
"back-flow" model. 4. Our results indicate that
the CDW in the sliding state interacts
differently with electrons and holes leading
strong change in the normal carriers
concentration at Et. Thanks to S.
Brazovskii and Yu. Latyshev